S&P Global's AI Customers Spend Far Faster, With API Calls Up More Than 4,000%
CEO Martina Cheung says AI-rights talks with clients are normalizing, as AI products add about half a point of growth in Energy and Kensho Data.
S&P Global customers who use its AI products are raising their spending much faster than everyone else, CEO Martina Cheung said on the company's third-quarter earnings call Tuesday, with API call volume up more than 4,000% from a year earlier. S&P Global, which provides benchmarks, data and analytics to capital, commodity and automotive markets, also said disputes over AI data rights are giving way to standard terms.
Usage up 50% in a quarter
Cheung said API call volume grew nearly 50% since the second quarter. More clients, she said, are building in-house tools and integrations with frontier AI models powered by S&P data.
In Market Intelligence, annual contract value growth among AI customers is 1.8 times that of non-AI customers. In Energy it is about three times, Cheung said. She expects the gap to close over time as more clients become AI customers.
Chief Financial Officer Eric Aboaf said discrete AI products added about 50 basis points (half a percentage point) to Energy revenue growth. They added about half a point to growth in Kensho Data and Platforms, a unit within Market Intelligence. Aboaf said the company had not described the effect in the first two quarters because it was still building.
AI-rights talks settle into standard terms
Cheung said negotiations over AI rights are "normalizing," and S&P is signing deals with robust pricing that protect its intellectual property. Sales cycles that stretched last quarter have stabilized.
Asked by a Wolfe Research analyst whether those terms are now standard, Cheung said clients in Energy and Market Intelligence have accepted what are essentially standard terms and conditions across the business. For adaptive retrieval agents, which run nuanced queries across several S&P datasets, she pointed to a potential for consumption-based pricing.
On cost savings, Aboaf declined to split AI out. He said the company budgets a couple of percentage points of productivity gains each year, and most of it mixes AI and non-AI changes.
Hyperscaler spending and the wider quarter
Cheung said S&P's team forecasts nearly $900 billion of capital spending this year by the six largest hyperscalers, nearly double the 2025 level.
Third-quarter revenue grew 7%, Aboaf said, and margin expanded 60 basis points to 53.5%. Adjusted EPS rose 9%. Management raised full-year adjusted EPS guidance to $17.70 to $17.95 and organic constant-currency revenue growth to 7.5% to 8.5%.
S&P also announced a strategic review of Capital IQ and associated data businesses, including a potential separation. Cheung said the earliest meaningful update would come on the February earnings call.
Cheung said she expects growth rates for AI and non-AI customers to converge over time as more clients adopt AI products.
Sources
- S&P Global Beats Earnings Estimates. A Major Shake-Up Could Be Coming. — Barron's
- S&P Global Schedules Third Quarter 2026 Earnings Announcement and Conference Call for Tuesday, October 27, 2026 - Sep 29, 2026 — press.spglobal.com
- S&P Global: Q3 Earnings Snapshot — Associated Press
- S&P Global Reports Third Quarter Results; Announces Strategic Review for Capital IQ and Associated Data Businesses - S&P Global — investor.spglobal.com
- S&P Global confirms strategic review of Capital IQ, data businesses; Q3 earnings beat (SPGI:NYSE) | Seeking Alpha — seekingalpha.com
